Density Peaking At Low Collisionality on Alcator C-Mod

Density Peaking At Low Collisionality on Alcator C-Mod APS-DPP Meeting Philadelphia, 10/31/2006 M. Greenwald, D. Ernst, A. Hubbard, J.W. Hughes, Y. Lin, J. Terry, S. Wukitch, K. Zhurovich, Alcator Group

Collisionality Effects on Density Profile ITER interest better fusion performance with moderate density peaking ASDEX-U H-Modes Results from ASDEX, JET suggest increase in density peaking at low ν* (note plots use ν eff = ν ei /ν De not ν*) Much of the previous work has significant beam fueling (C-Mod uses RF only) ASDEX-U: C. Angioni et al., PRL (2003) JET: H. Weisen, et al., NF (2005)

Collisionality Effects on Density Profile ITER interest better fusion performance with moderate density peaking ASDEX-U H-Modes Results from ASDEX, JET suggest increase in density peaking at low ν* (note plots use ν eff = ν ei /ν De not ν*) Much of the previous work has significant beam fueling (C-Mod uses RF only) Until recently, C-Mod density was too high to enter the very low ν* regime. Typical C-Mod H-modes

Low Density H-Modes Accessible With Unusual Shape With this shape, large ELMs appear in C-Mod for first time (Terry APS GI1.00004) n/n G = 0.3, ν eff = 0.33, ν ei τ E ~ 10

H-Mode In Standard C-Mod Shape EDA (QC mode) and small ELMs n/n G = 0.66, ν eff = 1.5, ν ei τ E ~ 40 (2x, 5x, 4x lower density H-mode respectively)

Lower Density H-Modes Show More Profile Peaking We don t know why these shapes allow operational H-mode at lower target densities Lower L-H threshold density ELMs seem to reduce rate of density rise in H-mode Whatever the cause: Standard H-mode n e0 /<n e > = 1.1-1.2 L n >> a, R/L n ~ 0, R/L T ~ 6-7 Low-Density H-mode n e0 /<n e > = 1.5 R/L n ~ 3-4, R/L T ~ 6-7

Difference In Profile Shape Is Notable Peaking is seen over outer 60% of plasma radius H-mode profiles evolve quickly; τ << a/v WARE

Difference In Profile Shape Is Notable Peaking is seen over outer 60% of plasma radius H-mode profiles evolve quickly; τ << a/v WARE Transport in center of plasma may be affected by sawteeth These are of large amplitude; δt e /T e ~ 25% Sawtooth Inversion Radius

Difference In Profile Shape Is Notable Peaking is seen over outer 60% of plasma radius H-mode profiles evolve quickly; τ << a/v WARE Transport in center of plasma may be affected by sawteeth These are of large amplitude; δt e /T e ~ 25% Radii chosen to characterize peaking in ASDEX-U are appropriate for C-Mod profiles r/a = 0.4 r/a = 0.8

Difference In Profile Shape Is Notable Peaking is seen over outer 60% of plasma radius H-mode profiles evolve quickly; τ << a/v WARE Transport in center of plasma may be affected by sawteeth These are of large amplitude; δt e /T e ~ 25% Radii chosen to characterize peaking in ASDEX-U are appropriate for C-Mod profiles r/a = 0.4 r/a = 0.8 So, collect a database and plot results.

Trend Of Increased Density Peaking At Low Collisionality Is Observed For ICRF Heated H-Modes In C-Mod

Trend Of Increased Density Peaking At Low Collisionality Is Observed For ICRF Heated H-Modes In C-Mod Standard Shape

Trend Of Increased Density Peaking At Low Collisionality Is Observed For ICRF Heated H-Modes In C-Mod New Shape Standard Shape

Trend Of Increased Density Peaking At Low Collisionality Is Observed For ICRF Heated H-Modes In C-Mod LHCD (L-Mode with V S ~ 0; V Ware very small) New Shape Standard Shape

C-Mod Data Overlays ASDEX-U Angioni PRL 2003 C-Mod

Same With Respect To JET Data. Weisen NF 2005 C-Mod

Same With Respect To JET Data. Weisen NF 2005 C-Mod ITER

n/n G Is Apparently Not A Good Scaling Parameter For Comparing C-Mod and JET Peaking Note horizontal separation of data sets C-Mod JET data from Weisen PPCF 2006 n/n G

n/n G Is Apparently Not A Good Scaling Parameter For Comparing C-Mod and JET Peaking Note horizontal separation of data sets ITER C-Mod JET data from Weisen PPCF 2006 n/n G

State Of Theory Is Not Yet Adequate To Explain These Results ITG/TEM turbulence can predict inward particle flux in the collisionless limit (Quasi-linear calculations: Angioni, Phys. Plasmas 12, 112310, 2005) Thermodiffusion mechanism (Coppi, PRL 41, 551,1978) Driven by T i, for R/L T > R/L n Pinch disappears at higher collisionality Depends on relative strength of ITG and TEM (L T, η, ν, T e /T i ) However: collisionality range with density peaking is predicted to be about 30 x smaller than what is seen in experiments Nonlinear GK simulations see pinch at experimentally realistic ν eff, but only by lowering L n /L T (Estrada-Milna, Phys. Plasmas 12 022305, 2005)

Summary Significant density profile peaking seen at low collisionality Trend and values match data from ASDEX-U and JET Lack of NBI fueling does not affect result Lack of Ware pinch does not affect result Result holds for Ti ~ Te Work extended to higher neutral opacity (ITER-like) Addition of C-Mod data suggest that ν eff is appropriate scaling variable rather than n/n G (These are strongly correlated, especially on any given machine) Good news for ITER Effect of shaping? Installation of cryopump and increased LH power will allow substantial extension of this work.